CpG island hypermethylation is a frequent event in late-stage ovarian cancer. At present, molecular information is needed for better stratification of this advanced disease in order to provide treatment options. The objective is therefore to identify this type of epigenetic biomarkers for predicting patients' survival and disease relapse. In a preliminary study, we employed a microarray-based approach, called differential methylation hybridization (DMH), for genome-wide screening in a small panel of stage Ill or lV tumors. Molecular profiling revealed a group of CpG island loci, the hypermethylation of which is significantly associated with shorter durations of progression-free survival in patients (p<0.001). These results have led us to hypothesize that a subset of hypermethylated loci are potential biomarkers for predicting treatment response in ovarian cancer. Our rationale is that increased methylation density in certain promoter CpG islands occurs in tumor progression and may be selected for during chemotherapy. This aberration may lead to the silencing of critical genes that regulate drug-detoxifying metabolisms and damage-induced signaling. As a result epigenetic dysregulation may contribute in part to the insensitivity of cancer cells to chemotherapy and to poorer outcome. In the R21 phase, we will conduct DMH for methylation screening in an extended panel of late-stage or relapse epithelial ovarian tumors. Quantitative milestones: a group of positive CpG island loci that exhibit statistical association with short survival and/or relapse will be identified and independently verified. In R33, a subpanel microarray of 25-50 such biomarkers will be established for efficient DMH screening in 300 epithelial ovarian tumors. This low-cost high-throughput method will critically evaluate whether hypermethylation of these loci correlate with patients advanced disease progression. Methylation-specific PCR assays of selected biomarkers will also be conducted in plasma DNA of ovarian cancer patients. Data generated in R33 will be used to establish clinical sensitivity and specificity of these methylation assays. Such a study lays the foundation for future prospective clinical trials designed to test the full utility of this microarray-based approach in routine clinical settings. Additionally, the study will give insights into the role of CpG island hypermethylation in drug-resistant ovarian cancer.